Fast Railway Clearance Inspection Using Mobile Laser Scanning Point Cloud

Railway clearance is to prevent the buildings and equipment of railway from intruding into the prescribed scope of railway.It is an important means of railway construction,safe operation and reconstruction of existing railway lines.Traditional railway clearance inspection has always been carried out by human pushing clearance box or visual inspection method,which has many procedures,long time and low efficiency.At the same time,the contact inspection in electrified road section also has a big safety risk.The non-contact 3D laser mobile scanner can quickly obtain the high precision and high density point cloud of railway,which is suitable for railway clearance inspection.At present,railway clearance inspection based on MLS point cloud mainly adopts cross-section method,which is time-consuming and low accuracy.1)A continuous railway clearance inspection algorithm based on clearance polyhedron is proposed.The invasion point cloud obtained by the topological relationship between the MLS point cloud and the clearance polyhedron.First of all,the railway clearance is initially detected by the axis-aligned bounding box of the standard clearance polyhedron.Then,through the preliminary inspection of the topological relationship between the possible invasion point and the clearance polyhedron,the invasion point cloud can be accurately detected,and the invasion distribution information can be obtained through the Euclidean clustering of the invasion point cloud.Taking MLS point clouds in railway tunnel,platform,and subgrade scenes as test data,the algorithm in this paper and the cross-section clearance inspection algorithm are respectively compared and tested.The results show that the accuracy of the algorithm in this paper in the inspection of invasion point cloud is 100%.Compared with the TIN based crosssectional clearance inspection algorithm,the recall rate of the proposed algorithm are improved 25.0%.Compared with the projection cross-section algorithm,the accuracy of the proposed algorithm are improved3.5%.In terms of time efficiency,the proposed algorithm is 81.7%higher than the cross-section algorithm.2)A railway clearance inspection algorithm based on MLS point cloud thinning algorithm is proposed.On the premise of retaining the shape feature point cloud of railway buildings or equipment,the improved curvature algorithm and the point cloud extraction algorithm are used to thin the railway point cloud,and then the clearance of the point cloud after the thinning is inspected.Then take the invasion point as the clustering center and cluster in the original point cloud.Then,principal component analysis is conducted on the point cloud after clustering and combined with the actual geographic information of railway buildings and equipment to get the invasion features.The clearance inspection was carried out on the original point cloud and the point cloud after thinning.The results show that the point cloud thinning can improve the railway clearance inspection efficiency,and the accuracy of the algorithm in this paper for the inspection of railway invasion features is 100%.